Seeking Tranquility in Busy Workplaces

# Brownian Noise, An Acoustic Firewall

Some thoughts on busy versus quiet work environments and why Brownian noise is a good choice to get some mentally challenging work done in a work place.

## Tranquility versus Turbulence

I am a guy who can enjoy silence. When working, there are times when I really need some silence. I have a tough time working on a task that requires my full concentration if there are environmental noises. I am talking of things

• employing a new non-trivial piece of technology (the first time I used boost::fusion was quite tough),
• or reading a math-heavy paper.

You may wonder why I am emphasizing this, as a calm working environment probably is a necessity for many people working in IT. As often in life, there are two kind of people and if you have only ever known your perspective, this may come as a surprise to you: some people need environmental noises. Typical bureau noises are a stimulant to them. Interestingly, in lab where I did my PhD, the guys working on the hardest math and theoretical CS problems were those who genuinely wanted some environmental business. I remember a scene during a conference where I preferred to stay in the hotel room to get some programming done, while my lab mate tried to write a paper on syntax trees and went into the hotel lobby for that.

So for everybody in my camp: how do we create a working environment where we thrive? Listening to engaging music (heavy metal, film music or epic music as produced by Two Steps from Hell) can be invigorating at times and can put you into that coffee-induced tunnel where you get quite some coding done (or at least you and your colleagues have the impression you are working heavily as you are typing a lot). But ultimately it is distracting when you are facing a mental challenge. And in particular, I am skeptic about whether the resulting product can be delivered quicker or whether one just produced a lot of half-baked things that need refactoring later.

Music without lyrics is better, but there is still too much going on. I have tried out rain or beach sounds, but I came to the conclusion what you ultimately want is something completely constant. Otherwise distractions sneak in through the holes when there is less rain or beach waves.

## Enter the Noise

This is how I finally ended up listening to…well…noise. Because noise in itself is not the problem, instead it is the patterns in the noise which are perceived as a distraction. It is relatively easy to blend out a constant sound source (well, this naturally does not hold true for Tinitus and the like). But noise does not equal noise, there are many flavours. While my background in statistics enabled me to understand some analogies to statistical processes and distributions, my knowledge of physics and acoustics is rather limited. So dear physicists please have mercy when going through my understanding of the colors of noise:

• What the lay man commonly talks about when referring to noise is white noise which is what you may know as the static in a TV or radio without signal. Effectively, it means that all frequencies (i.e., high and low tones) have an equal amplitude (i.e., are equally loud). Interestingly, we perceive white noise as somewhat high-pitched as we perceive different frequencies with equal amplitude as being differently loud. Okay, this should not surprise any human living past 1900: I hope, it does not shatter your world view, but we do not perceive reality objectively (assuming an objective reality exists).

• The German version of Wikipedia claims that this is different for pink noise or $\frac{1}{f}$ noise. Pink noise means that higher tones become less loud or in other words, per octave (frequency) we get 3 dB less loud. As we only care about orders of magnitude for both frequency as well as amplitude, this means that we have a linear relationship in a double logarithmic plot characterizing pink noise. The analogous plot for white noise depicts a constant. We would expect that our hearing is optimized for a certain frequency spectrum and is suboptimal on both tails of this distribution. The type of noise that compensates for our acoustic bias is called grey noise and therefore I find it highly unlikely, that pink noise means that we perceive all frequency as being equally loud.

• Now here is the insider tip and the entire point of the article: Brownian noise or $\frac{1}{f^2}$ noise. Essentially, it is a version of pink noise that favours deeper frequencies even more (i.e., the loudness decreases by 6 dB per octave). From my experience, Brownian noise is optimal to blend out any environmental patterns at a minimum overall loudness of noise. Of course, you could turn up white noise to a level such that it overshadows any distracting sound, but listening to loud noise for a long time is surely not optimal for your eardrum. I may be biased a bit, since I work in a business with predominantly male coworkers and male voices are deeper on average. But that being said, deep noise in my eyes (eeeh, I meant ears) are more calming than more high-pitched noise. I have listened to samples of “super-deep” Brownian noise, which are relaxing but they do not cover higher environmental sounds well.

So there we have our conclusion: I have gone for Brownian noise and did not get bored with it so far. Well, okay, the entire point is that you forget that you are listening to it.

A little side note: I was able to test noise cancelling headphones once. These headphones capture environmental sounds via microphone and produce sound which cancels it out. This works well to with sound remaining constant, but it works notoriously poorly with modulating signals like human voices. So if it is human speech that distracts you, noise cancelling headphones actually make it worse. Save your money on that and directly go for Brownian noise.